Venky Ramesh is Chief Client Officer at LatentView Analytics, with two decades of experience leading businesses and teams globally.
One of the most useful business books I’ve read is The Culture Map by Erin Meyer. I first picked it up to become a better leader for global teams. I didn’t expect it to change how I think about AI.
The book explains something every international executive eventually learns: People don’t just speak different languages—they think, communicate and make decisions differently. Americans tend to communicate directly. Japanese communication relies much more on context. Germans are comfortable separating disagreement from relationships. In many Asian cultures, trust is built before business, whereas Americans often build trust by doing business together. None of these approaches is better than another. They’re simply different operating systems.
As AI becomes another participant in how work gets done, a question emerges: Can it understand the cultural context that humans navigate every day? We spend countless hours discussing models, reasoning, memory and agents. Yet, one of the biggest variables in AI success may be something much older than technology: culture.
Same Process, Different Conversations
Over the last two decades, I’ve worked with teams across the U.S., India, Mexico and Eastern Europe. The same business process can produce very different conversations. In the U.S., meetings tend to get to the point quickly and decisions are often made in the room. In India, the real meeting sometimes begins after the meeting, as people build alignment before execution. In Mexico, relationships often influence how business gets done. In Eastern Europe, I’ve found teams remarkably analytical and comfortable challenging ideas if it leads to a better answer.
Every individual is different, but these patterns reflect something The Culture Map captures beautifully: Culture shapes how work gets done. People learn to recognize and adapt to these different operating systems. AI will need to do the same. Today, it largely learns from policies, workflows and documentation. The problem is that much of what makes global companies work never appears in a policy document.
Take something as simple as approving an invoice. In many U.S. companies, the rule might be straightforward: Approve anything below $5,000 unless it violates procurement policy. An AI agent can check the amount, validate the policy and move on. Now, place that same agent in a high-context culture. The policy still says $5,000, but someone points out that the supplier stood by the company during a shortage three years ago. Another mentions that this customer is strategically important. A senior leader says, “Technically, the policy says no, but we’ve handled these situations differently before.”
Every person in the room understands exactly what’s happening. The AI is still searching for the missing page of the policy manual. That’s the difference between low-context and high-context cultures. In one, most of the information is explicit. In the other, much of it lives in relationships, shared history and unwritten norms.
Global Policies, Local Context
This becomes especially important for global companies. A consumer goods company may have headquarters in New York, manufacturing in Mexico, sourcing in India, commercial teams in Brazil and customers across Europe and Asia. The policies may be global, but the context is always local. An AI system that performs brilliantly in one culture may struggle in another, not because it’s less intelligent but because it’s applying the wrong cultural operating system.
Conclusion
The future of enterprise AI isn’t just multilingual—it’s multicultural. It’s AI that knows when direct feedback is appreciated and when it’s seen as disrespectful. It’s AI that understands when consensus is part of making the decision rather than delaying it. And it’s AI that recognizes that the technically correct answer isn’t always the culturally effective one.
Data gives AI information. Reasoning gives AI answers. Culture gives those answers context. In the end, enterprise AI won’t be measured by how many decisions it makes but by how well those decisions work across different people, markets and cultures, ultimately reflected in adoption and results.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
